Cannabinoid action in the olfactory epitheliumDirk Czesnik*†, Detlev Schild*‡, Josko Kuduz*, and Ivan Manzini*‡

*Department of Neurophysiology and Cellular Biophysics and ‡Deutsche Forschungsgemeinschaft Research Center for Molecular Physiology of the Brain,University of Göttingen, Humboldtallee 23, 37073 Göttingen, Germany

(Edited by L. L. Iversen, University of Oxford, Oxford, United Kingdom, and approved December 20, 2006 (received for review October 13, 2006)

The perception of odors is influenced by a variety of neuromodu-lators, and there is growing evidence that modulation alreadytakes place in the olfactory epithelium. Here we report on canna-binergic actions in the olfactory epithelium of Xenopus laevistadpoles. First we show that CB1 receptor-specific antagonistsAM251, AM281, and LY320135 modulate odor-evoked calciumchanges in olfactory receptor neurons. Second, we localize CB1-likeimmunoreactivity on dendrites of olfactory receptor neurons. Fi-nally, we describe the cannabinergic influence on odor-inducedspike-associated currents in individual olfactory receptor neurons.Here we demonstrate that the cannabinergic system has a pro-found impact on peripheral odor processing and discuss its possiblefunction.

CB1 antagonist � dendrite � modulation � odor processing �olfactory receptor neuron

Sensory olfactory information is crucial in several behavioralaspects of humans and animals, e.g., nutrition and repro-duction. It is everyday experience that olfactory stimuli that areattractive before food intake may become neutral or evenaversive after. Functional MRI in humans showed that theactivation of an orbitofrontal cortex area upon stimulation withbanana odor was diminished after bananas were eaten to satiety(1). At a more peripheral stage, in mitral cells of the rat olfactorybulb, the responsiveness to food odors decreased when rats werefed, whereas it increased after fasting (2, 3). Some modulationmay even occur in the olfactory epithelium (OE). In hungryaxolotls s-neuropeptide Y enhances EOG responses to L-glutamic acid and modulates the amplitude of a tetrodotoxin-sensitive inward current (4). S-neuropeptide Y generally appearsto play an important role in the control of appetite and feeding(5, 6).

Recently, the endocannabinergic system (ECS) has also beenshown to be involved in food intake and energy homeostasis (7).Because in different animal phyla the levels of endocannabinoidsare increased under fasting conditions (8–11), endocannabinoidsmay act as orexigenic mediators. These observations, togetherwith the role of olfaction in food detection, led us to investigatewhether olfactory receptor neurons (ORNs) are modulated bythe cannabinergic system, which clearly turned out to be the case.

ResultsCB1-Specific Antagonists Reduce Odor-Evoked Calcium Changes inORNs. A mixture of 19 amino acids (100 �M, AAMIX) appliedto an acute slice preparation of the OE of Xenopus laevis tadpoles(12) led to a specific cellular response pattern (Fig. 1 A and B)(13, 14). Amino acids are well known olfactory stimuli in aquaticanimals involved in feeding behavior (15–17), and they areknown to activate a subset of ORNs in the OE of larval X. laevis(13, 14). The response time courses of four amino acid-sensitiveORNs are plotted in Fig. 1C. Shape and duration of theintracellular calcium transients were highly reproducible whenthe odorants were applied repeatedly.

After the identification of amino acid-sensitive ORNs weadded the cannabinergic antagonist AM251 to the bath solution.After drug wash-in the latency of the response to AAMIX wasmarkedly increased (Fig. 1D; AM251, 5 �M, red curves). Theamplitude as well as the delay and the duration of the responses

were modulated in a stereotypic way. The amplitudes decreased,whereas the delay and the duration of the responses increased.After 12 min of washout the odor responses recovered (greencurves). The degree of the described effects varied from ORNto ORN (Fig. 1D). The CB1 antagonists AM281 and LY320135gave similar results (data not shown). The above-describedmodulation was observed in a total of 182 ORNs (18 slices from18 different animals). The antagonists were used at concentra-tions of 1, 5, 10, and 20 �M. After a washout period of �15 min,8 of 8 ORNs (1 �M), 73 of 79 ORNs (5 �M), and 19 of 56 ORNs(10 �M) recovered. With 20 �M antagonist concentration noneof the 39 ORNs showed a recovery.

Application of AM251/AM281 alone did not alter the [Ca2�]ilevel (15 slices, concentration range from 1 �M to 100 �M) (Fig.1E1). In addition, the coapplication of amino acids and AM251did not change the amino acid response (Fig. 1 E2 and E3). Thisexcludes a direct effect of the antagonists on the olfactoryreceptor proteins.

Furthermore, we excluded the possibility of non-CB1-mediated actions of the antagonists used. We were able to showthat the recovery time after AM281 (10 �M) application couldbe markedly shortened by addition of the highly specific CB1receptor agonist HU210. At an antagonist concentration of 10�M a recovery from the antagonist effect was never seen before5 min of washout with bath solution alone (34 ORNs in threeslices) (Fig. 2A), whereas addition of HU210 (20 �M) led to analmost complete recovery of amino acid responses after 2 min ofwashout (Fig. 2B). Similar results were obtained in 36 cells ofthree slices. In addition, HU210 increased the statistical fre-quency of recovering responses from 19 of 56 (33%) ORNs (fiveslices) to 36 of 36 (100%) ORNs (three slices).

Distribution of CB1 Receptors in the OE. Using immunohistochem-istry we found CB1 receptor-like immunoreactivity (CB1-LI IR)in the OE of X. laevis tadpoles (anti-rat CB1 N terminusantibody, kindly gifted by K. Mackie, University of Washington,Seattle, WA). ORNs were visualized by biocytin fills of theolfactory nerve (see Materials and Methods). Fig. 3A shows thetypical localization and shape of ORNs. The somata of thebiocytin-avidin-stained cells are clearly located in the ORNlayer.

CB1-LI IR is localized to dendritic structures mainly in theapical part of the OE (Fig. 3B). The white rectangle in Fig. 3Aindicates the region shown at higher magnification in Fig. 3 D–F.Dendrites and somata of individual ORNs can be clearly iden-tified, and the CB1-LI IR is clearly associated with the dendriticcompartments (Fig. 3 E and F). Somata and axons show noimmunostaining (Fig. 3A). The specificity of CB1-LI IR wasconfirmed by the absence of immunostaining in sections treated

Author contributions: D.C., D.S., and I.M. designed research; D.C., J.K., and I.M. performedresearch; D.C. and I.M. analyzed data; and D.C., D.S., and I.M. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS direct submission.

Abbreviations: OE, olfactory epithelium; ECS, endocannabinergic system; ORN, olfactoryreceptor neuron; AAMIX, mixture of amino acids; CB1-LI IR, CB1-like immunoreactivity.

†To whom correspondence should be addressed. E-mail: dczesni@gwdg.de.

© 2007 by The National Academy of Sciences of the USA

www.pnas.org�cgi�doi�10.1073�pnas.0609067104 PNAS � February 20, 2007 � vol. 104 � no. 8 � 2967–2972

NEU

ROSC

IEN

CE

Dow

nloa

ded

by g

uest

on

June

14,

202

1

with the anti-CB1 antibody preadsorbed with the immunizingfusion protein (Fig. 3C, as compared with Fig. 3B).

AM251 Alters Odor-Evoked On-Cell Spikes. The observed odorant-induced [Ca2�]i changes in ORNs allow only indirect conclusionsregarding the information conveyed to the olfactory bulb. Wetherefore tested the effect of the CB1-specific antagonists onodor-induced spiking activity by recording spike-associated cur-rents of individual ORNs. Fig. 4A1 shows a typical odor responseof an individual ORN measured in the on-cell patch-clampconfiguration. The poststimulus time histogram of the associatedcurrents allows a comparison of repeated odorant responses(Fig. 4B1; first application, black curve; second application, redcurve).

The application of AM251 (5 �M) increased the delay in theonset of the odor-induced spike-associated currents and theinterspike interval of the responses (Fig. 4 A2–A4 and B2–B4)although the stimulus was present at the OE for several seconds(Fig. 4A6). Similar results were obtained in seven other ORNs(n � 7 slices). Washout of the antagonists took some time so thata recovery could be observed only after �15 min (Fig. 4 A5 andB5). Fig. 4C shows individual spike-associated currents at highertime resolution, each taken from the trace indicated. Amplitude

and duration of the individual currents were not affected by theantagonists.

Taken together, odor responses under CB1 antagonists areincreasingly delayed with drug wash-in. While they are graduallymore and more inhibited, they become longer-lasting and lessintense (in terms of spikes per second). The effects of the CB1antagonists on spiking activity are thus consistent with theeffects on [Ca2�]i above.

DiscussionIn the present study we show the influence of the ECS on odorprocessing in the OE of X. laevis tadpoles. By means of confocal[Ca2�]i imaging, patch clamp recordings, and immunohisto-chemistry we were able to locate the CB1 receptors and todescribe the effects of CB1-specific antagonists.

In an acute slice preparation of the OE of X. laevis tadpoleswe found that in ORNs the specific CB1 receptor antagonistsAM251, AM281, and LY320135 decrease the amplitude ofodor-evoked [Ca2�]i traces and increase the latency to activation.We can exclude the possibility of a competitive antagonism ofthese drugs at the olfactory receptor proteins because (i) theantagonists showed the same effect, (ii) applying the drugs alonedid not induce a response, and (iii) coapplication of a mixture ofboth amino acids and a high antagonist concentration neveraltered the amino acid-induced response.

Furthermore, we excluded any non-CB1-mediated effect ofthe used antagonists. We found that the highly specific CB1agonist HU210 drastically accelerates the recovery during wash-out and increases the percentage of recovering responses.

CB1-LI IR is distributed mainly in the apical layer of the OEof X. laevis tadpoles, where generally both the dendrites of ORNsand the cell bodies of sustentacular cells are located (Fig. 3 A andB). Higher magnification together with visualization of ORNmorphology by double staining with biocytin-avidin Alexa Fluor488 allowed us to attribute the CB1-LI IR to ORN dendrites.ORN dendrites, which are present in all ORNs, from snails tohumans, are certainly the appropriate compartment for partiallydecoupling the transduction compartment from the transforma-tion compartment, and endocannabinoids may play a modula-tory role in this compartment. Our evidence that CB1 receptorsare located on the dendritic compartment and that CB1 antag-onist application decreased responsiveness to odors is consistentwith such a speculation.

Several studies show that CB1 receptors or the related mRNAoccur at different stages of the central olfactory system in variousanimal phyla (18, 19). Our study extends this observation to theOE, the first stage of the olfactory system. Recent evidence

Fig. 1. AM251 alters odor-evoked [Ca2�]i changes. (A) Overview of a X. laevistadpole head. The black rectangle indicates part of the animal’s left OE. (Scalebar: 500 �m.) (B) Fluo4-AM stained acute slice preparation of the OE (imageacquired at rest). PC, principal cavity. The yellow ovals indicate the ORNsomata of this slice that responded to the AAMIX (100 �M). (Scale bar: 10 �m.)(C) [Ca2�]i transients of individual ORNs upon repeated applications of theAAMIX. The intraepithelial location of the four cells shown is indicated in B.[Scale bars: 10 s and �F/F 100% (cells 1 and 2) and 10 s and �F/F 50% (cell 3 andcell 4).] (D) After addition of AM251 (5 �M) to the bath solution the AAMIX-evoked ORN responses (black traces) were modulated (red traces). After 12min of drug washout the odor-induced [Ca2�]i transients recovered com-pletely (green traces). [Scale bars: 10 s and �F/F 100% (cells 1 and 2) and 10 sand �F/F 50% (cells 3 and 4).] (E) AM251 does not interfere with odorantbinding at olfactory receptors. Application of AM251 (50 �M) alone (E1) didnot elicit any response. (E2 and E3) Odorant-induced [Ca2�]i transient uponapplication of the AAMIX (50 �M) alone (E2) and [Ca2�]i transient induced bya coapplication of the antagonist (AM251, 50 �M) and 50 �M AAMIX (E3). Thereproducibility of the [Ca2�]i transients was not altered by the presence of theantagonist (compare black and red traces). (Scale bars: 10 s and �F/F 100%.)

Fig. 2. CB1 receptor-mediated antagonist action upon amino acid-sensitiveORNs. (A) The modulatory effects of AM281 (10 �M; red trace) on an AAMIX-evoked ORN response (black trace) could not be washed out within 2 or 5 minwith bath solution alone (green trace and blue trace, respectively). (B) [Ca2�]itransient of an individual ORN upon application of the AAMIX (black trace).After addition of AM281 (10 �M) to the bath solution the AAMIX-evoked ORNresponse was suppressed (red trace). After 2 min of drug washout with HU210(20 �M) in the bath solution, the odor-induced [Ca2�]i transient recoveredalmost completely (green trace). [Scale bars: 10 s and �F/F 50% (A) and 10 s and�F/F 100% (B).]

2968 � www.pnas.org�cgi�doi�10.1073�pnas.0609067104 Czesnik et al.

Dow

nloa

ded

by g

uest

on

June

14,

202

1

indicated the presence of CB1 mRNA in OE of X. laevis tadpolesat stage 46 (20). In our study we now describe the subcellulardistribution of CB1 receptors in the OE of larval X. laevis. Moreimportantly, we present the first physiological data of cannab-inergic actions in ORNs upon odorant stimulation. Photondetection in the retina also seems to be modulated by the ECS(21–26). In addition, endocannabinergic signaling may also existbetween taste cells and taste nerves (27). The endocannabinergicmodulation of sensory output at the most peripheral stage maythus be a common feature of at least some sensory systems.

[Ca2�]i imaging is a well established tool to reveal the sensi-tivity to odors (28), but it gives no direct information regardingwhether the signal is relayed to higher brain centers. This can beconcluded unambiguously only by directly measuring electricalactivity of individual cells. We therefore recorded odor-inducedtrains of spike-associated currents. The observed modulatoryeffect by AM251 is comparable to the modulation detected inour [Ca2�]i imaging experiments. Spiking responses were in-creasingly delayed, and they became longer and weaker, al-though the extent to which these effects occurred varied fromORN to ORN. Future experiments may show the relativecontributions of the antagonist effects on response strength, itsdelay, and its length. In any case, it can already be stated withcertainty that the cannabinergic system drastically alters theafferent input to the olfactory bulb.

AM251, AM281, and LY320135 act as both CB1-specificantagonists and inverse agonists (29). It has therefore to beclarified whether the observed modulation is due to a compet-itive antagonism or an inverse agonism. A tonic release ofendocannabinergic substances may be assumed, because CB1agonist application alone during repetitive odor applicationusually showed little or no effect (data not shown). Thus, thecontrol of the endocannabinergic modulation of odorant-induced responses in ORNs needs to be investigated in moredetail in future experiments. First steps in this direction wouldbe a quantification of endocannabinoid levels in the OE com-bined with a detailed characterization of the involved mecha-nisms regarding re-uptake and enzymatic hydrolysis. Our findingthat the ECS influences olfactory sensory input adds to a

growing knowledge showing that the activity of ORNs is mod-ulated by numerous compounds, including acetylcholine, adren-aline, ATP, dopamine, GnRH, and s-neuropeptide Y (4, 30–36).

Recently, several studies have been published dealing with theinfluence of the nutritious status on the neurophysiology ofolfactory information processing and vice versa, whereby someof the phenomena could indirectly be attributed to the effects ofmodulators like orexin in the rat olfactory bulb or s-neuropeptide Y in the OE (4, 37–39). The ECS is also involvedin food intake and energy homeostasis (7). For instance, in theteleost fish Carassius auratus (8), in the zebra finch (9), and inrodents (10, 11), brain endocannabinoids seem to act as orexi-genic mediators. In addition, AM251 induces suppression of ratfood intake and food-reinforced behavior (40). Thus, our find-ings together with the above-mentioned observations and theknown role of olfaction in food detection support the view thatthe ECS may play an important role in the response of organismsto their nutritional status, which has to be clarified in futurestudies.

Materials and MethodsSlice Preparation for Calcium Imaging and Patch Clamping. Tadpolesof X. laevis (stages 51–54) (41) were chilled in a mixture of iceand water and decapitated, as approved by the University ofGöttingen Committee for Ethics in Animal Experimentation. Ablock of tissue containing the OE, the olfactory nerves, and thebrain was cut out and kept in bath solution (see below). Thetissue was then glued onto the stage of a vibroslicer (VT 1000S;Leica, Bensheim, Germany) and cut horizontally into 130- to150-�m-thick slices. For patch clamping the slices were placedunder a grid in a recording chamber and viewed by usingNomarski optics (Axioskop 2; Zeiss, Göttingen, Germany). Forimaging [Ca2�]i the tissue slices were incubated with 200 �l of abath solution (see below) containing 50 �M fluo-4 AM (Mo-lecular Probes, Leiden, The Netherlands) and 50 �M MK571(Alexis Biochemicals, Grünberg, Germany). Fluo-4 AM wasdissolved in DMSO (Sigma, Deisenhofen, Germany) and Plu-ronic F-127 (Molecular Probes). The final concentrations ofDMSO and Pluronic F-127 did not exceed 0.5% and 0.1%,

Fig. 3. CB1 distribution in the OE of X. laevis tadpoles. (A) Slice of an OE with biocytin-avidin-stained ORNs (merged z-stack; 16 optical slices, total thickness14.4 �m). (Scale bar: 20 �m.) (B) Immunoreactivity to an rCB1-NH antibody of the same slice (see Materials and Methods). (C) Merged z-stack of an OE treatedwith the anti-CB1 antibody after preadsorption with the immunizing protein (see Materials and Methods). (D–F) Higher magnification of the region indicatedby the white rectangle in A (merged z-stack; four optical slices, total thickness 3.6 �m). (Scale bar: 10 �m.) (E) CB1-LI IR localized to dendritic processes. The somataof the ORNs are clearly IR-free. (F) Overlay of D and E. CB1-LI IR is clearly occurring at ORN dendrites.

Czesnik et al. PNAS � February 20, 2007 � vol. 104 � no. 8 � 2969

NEU

ROSC

IEN

CE

Dow

nloa

ded

by g

uest

on

June

14,

202

1

respectively. To avoid multidrug resistance transporter-mediated destaining of the slices, MK571, a specific inhibitor ofthe multidrug resistance-associated proteins, was added to theincubation solution (42). After incubation at room temperaturefor 35 min, the tissue slices were put under a grid in a recordingchamber and placed on the microscope stage of an Axiovert100M (Zeiss, Jena, Germany) to which a laser scanning unit(LSM 510; Zeiss) was attached. Before starting the calcium-imaging experiments, the slices were rinsed with bath solutionfor at least 15 min.

Calcium Imaging of Odor Responses. [Ca2�]i was monitored by usinga laser-scanning confocal microscope (LSM 510). The confocalpinhole was set to �140 �m to exclude fluorescence detectionfrom more than one cell layer. Fluorescence images (excitationat 488 nm; emission � 505 nm) of the OE were acquired at 1.27Hz, with three to five images taken as control images before theonset of odor delivery. The fluorescence changes �F/F werecalculated for individual ORNs as �F/F � (F1 � F2)/F2, whereF1 was the fluorescence averaged over the pixels of an ORN

soma and F2 was the average fluorescence of the same pixelsbefore stimulus application, averaged over three images. Aresponse was assumed if the following two criteria were met: (i)the first two intensity values after stimulus arrival at the mucosa,�F/F(t1) and �F/F(t2), had to be larger than the maximum of theprestimulus intensities; and (ii) �F/F(t2) � �F/F(t1) with t2 � t1.Data analysis was performed with Matlab (Mathworks).

Patch Clamp Recordings. Patch electrodes with a tip diameter of1–2 �m and �7–10 M� resistance were fabricated from boro-silicate glass with a 1.8-mm outer diameter (Hilgenberg, Mals-feld, Germany) by using a two-stage electrode puller (Narishige,Tokyo, Japan). ‘‘On-cell configuration’’ recording of a cell wasdone after a gigaohm seal was obtained between the patchpipette and the membrane of an individual intact cell. Thisnoninvasive technique makes it possible to record action poten-tial-equivalent charge displacements of the membrane of anindividual cell without affecting the composition of its intracel-lular solution (43). Holding voltage was 0 mV. Pulse protocols,data acquisition, and evaluation programs were written in C.

Fig. 4. CB1 antagonist AM251 alters odor-evoked electrical activity in individual ORNs. (A1) AAMIX-induced action potential-associated currents of anindividual ORN. [Scale bars: 2 s and 20 pA (A1–A5).] The corresponding poststimulus time histogram is shown in B1. The superimposed red histogram comes froma successive AAMIX application (interstimulus interval 3 min; current trace not shown) and shows the high reproducibility of the odorant response. (A2–A4 andB2–B4) The modulatory effect of AM251 on the action potential-associated currents depends on the wash-in time of the antagonist. (A5 and B5) Recovery after20-min drug washout. The time window of the original response is indicated by the gray-shaded area. (A6 and B6) Odor concentration time course. Shown issimulation of the odor dynamics during a single odorant application (see Materials and Methods). (C1–C5) Zoom of individual spike-associated currents, one fromeach current trace A1–A5 (trace duration 25 ms). (Scale bar: 40 pA.)

2970 � www.pnas.org�cgi�doi�10.1073�pnas.0609067104 Czesnik et al.

Dow

nloa

ded

by g

uest

on

June

14,

202

1

Voltage pulses were delivered from a microcontroller to a D/Aconverter and then to the patch clamp amplifier (EPC7; List,Darmstadt, Germany). The data were digitized online. Furtherdata analysis was performed with Matlab.

Solution and Stimulus Application. The composition of the bathsolution was 98 mM NaCl/2 mM KCl/1 mM CaCl2/2 mM MgCl2/5mM glucose/5 mM sodium pyruvate/10 mM Hepes. The bathsolution was also used as pipette solution for the on-cell record-ings. The pH was adjusted to 7.8. The osmolarity of the solutionwas 230 mOsmol/liter. As odorants, we used the same mixture of19 amino acids as in our previous work (12). The amino acidswere dissolved in bath solution (stocks of 10 mM) and used at afinal concentration of 50–100 �M in all of the experiments.Stimulus solutions were prepared immediately before use bydissolving the respective stock solution in bath solution. The bathsolution was applied by gravity feed from a storage syringethrough a funnel drug applicator to the recording chamber.Stimuli were pipetted directly into the funnel without stoppingthe flow. Outflow was through a syringe needle placed close tothe OE. The time course of stimulus arrival at the OE wassimulated by applying the fluorescent dye avidin Alexa Fluor 488as a dummy stimulus and by measuring the fluorescence afteravidin Alexa Fluor 488 application to the funnel. The delay ofstimulus arrival caused by the syringe, i.e., from pipetting into thefunnel to concentration increase in the OE, was �2 s (see Fig.4A). The minimum interstimulus interval between odorantapplications was at least 2 min. All of the chemicals werepurchased from Sigma if not otherwise indicated. The CB1receptor drugs AM251, AM281, LY320135, and HU210 (Tocris,Bristol, U.K.; stocks of 10 mM or 20 mM, 100% DMSO) weredissolved in bath solution and used at final concentrationsindicated in Results or the figure legends.

Immunohistochemistry. ORNs of X. laevis tadpoles (stages 51–54)(41) were backfilled with biocytin (Molecular Probes) throughthe olfactory nerve. A crystal of biocytin was put into the nerveof anesthetized tadpoles. After the lesion was closed withhistoacryl glue, the tadpoles were put back into the aquarium.Two hours later the tadpoles were chilled again and a block oftissue containing the OE was cut out, transferred into a solutionof 4% formaldehyde in PBS (pH 7.4), and fixed for 2 h at roomtemperature. The tissue was then washed at least three times in

PBS, embedded in 5% low-melting-point agarose (Agarose typeII; Amreso, Solon, OH), and sectioned at �70 �m on a vibro-slicer (VT 1000S). The slices were then immersed in a solutionof avidin Alexa Fluor 488 conjugate (5 mg/ml; Molecular Probes)in PBS-TX (0.2%) overnight.

The OE slices were then double-stained with an affinity-purified primary polyclonal antibody raised against the N ter-minus of the rat CB1 receptor (rCB1-NH, raised in rabbit, kindlyprovided by Ken Mackie). We obtained the best results whensections were preincubated (1 h at room temperature) inTBS-TX and 2% normal goat serum (NGS) (ICN Biomedicals,Orsay, France). Background was reduced after preincubation inNGS. The tissue was then incubated overnight at 4°C withrCB1-NH (1:500) diluted in TBS-TX with 2% NGS. Sectionswere subsequently washed with TBS, and Alexa Fluor 546-conjugated anti-rabbit secondary antibody (Molecular Probes)was applied at a dilution of 1:250 in 1% NGS/TBS for 2 h at roomtemperature. The secondary antibody was washed off by fivechanges of TBS. The preparations were then transferred into60% glycerol/PBS for at least 1 h and finally mounted on slidesfor confocal microscopy in 80% glycerol/PBS.

The specificity of the anti-CB1 antibody was previously as-sessed in rat (44) and evaluated in X. laevis tadpoles by incu-bating the sections with the anti-CB1 antibody (1:500) pread-sorbed (1 h at room temperature) with the immunizing fusionprotein (1 �g). Preparations were viewed by using a laser-scanning confocal microscope attached to an inverted micro-scope (LSM 510). Series of optical sections were imaged atintervals of 0.9 �m through the depth of the thick sections. Theywere saved as single optical images or three-dimensional stacks.Two-dimensional projections were generated for each channeland merged with the use of pseudocolors. Image processing wasperformed by using either Zeiss imaging software or GIMP(GNU Image Manipulation Program, www.gimp.org).

We thank Gudrun Federkeil for excellent technical assistance and Dr.Ken Mackie for generous supply of antibodies. This work was supportedby grants from the Research Program of the Faculty of Medicine ofGeorg-August-Universität Göttingen (to D.C.) and the Deutsche For-schungsgemeinschaft Research Center for Molecular Physiology of theBrain (to D.S. and I.M.). Generation and purification of the rCB1-NHantibody was supported by the National Institutes of Health GrantDA11322.

1. O’Doherty J, Rolls ET, Francis S, Bowtell R, McGlone F, Kobal G, Renner B,Ahne G (2000) NeuroReport 11:399–403.

2. Chaput M, Holley A (1976) Chem Senses 2:198–201.3. Pager J, Giachetti I, Holley A, Le Magnen J (1972) Physiol Behav 9:573–

579.4. Mousley A, Polese G, Marks NJ, Eisthen HL (2006) J Neurosci 26:7707–7717.5. Kalra SP, Kalra PS (2004) in Neuropeptide Y and Related Peptides, ed Michel

MC (Springer, Berlin), pp 221–249.6. Levens NR, Feletou M, Galizzi JP, Fauchere JL, Della Zuana O, Lonchamt M

(2004) in Neuropeptide Y and Related Peptides, ed Michel MC (Springer, Berlin),pp 283–235.

7. Matias I, Di Marzo V (2005) Nat Neurosci 8:585–589.8. Valenti M, Cottone E, Martinez R, De Pedro N, Rubio M, Viveros MP,

Franzoni MF, Delgado MJ, Di Marzo V (2005) J Neurochem 95:662–672.9. Soderstrom K, Tian Q, Valenti M, Di Marzo V (2004) J Neurosci 24:10013–

10021.10. Kirkham TC, Williams C, Fezza F, Di Marzo V (2002) Br J Pharmacol

136:550–557.11. Di Marzo V, Goparaju SK, Wang L, Liu J, Batkai S, Jarai Z, Fezza F, Miura

GI, Palmiter RD, Sugiura T, Kunos G (2001) Nature 410:763–765.12. Manzini I, Rössler W, Schild D (2002) J Physiol 545:475–484.13. Manzini I, Schild D (2003) J Physiol 551:115–123.14. Czesnik D, Kuduz J, Schild D, Manzini I (2006) Eur J Neurosci 23:119–128.15. Sorensen, Caprio (1998) in The Physiology of Fishes, ed Evans DH (CRC, Boca

Raton, FL), pp 251–261.16. Valentincic T, Lamb CF, Caprio J (1999) Physiol Behav 67:567–572.

17. Nikonov AA, Finger TE, Caprio J (2005) Proc Natl Acad Sci USA 102:18688–18693.

18. Cesa R, Mackie K, Beltramo M, Franzoni MF (2001) Cell Tissue Res 306:391–398.

19. Egertova M, Elphick MR (2000) J Comp Neurol 422:159–171.20. Beatrice M, Gabriella M, Francesca G, Oliana C (2006) FEBS Lett 580:1941–

1945.21. Straiker A, Stella N, Piomelli D, Mackie K, Karten HJ, Maguire G (1999) Proc

Natl Acad Sci USA 96:14565–14570.22. Straiker A, Sullivan JM (2003) J Neurophysiol 89:2647–2654.23. Fan SF, Yazulla S (2003) Vis Neurosci 20:177–188.24. Fan SF, Yazulla S (2004) Vis Neurosci 21:69–77.25. Fan SF, Yazulla S (2005) Vis Neurosci 22:55–63.26. Struik ML, Yazulla S, Kamermans M (2006) Vis Neurosci 23:285–293.27. Oike H, Wakamori M, Mori Y, Nakanishi H, Taguchi R, Misaka T, Matsumoto

I, Abe K (2006) Biochim Biophys Acta 1761:1078–1084.28. Manzini I, Schild D (2004) J Gen Physiol 123:99–107.29. Pertwee RG (2005) Life Sci 76:1307–1324.30. Arechiga C, Alcocer H (1969) Exp Med Surg 27:384–394.31. Bouvet JF, Delaleu JC, Holley A (1988) Neurosci Res 5:214–223.32. Kawai F, Kurahashi T, Kaneko A (1999) Nat Neurosci 2:133–138.33. Vargas G, Lucero MT (1999) J Neurophysiol 81:149–158.34. Eisthen HL, Delay RJ, Wirsig-Wiechmann CR, Dionne VE (2000) J Neurosci

20:3947–3955.35. Hegg CC, Greenwood D, Huang W, Han P, Lucero MT (2003) J Neurosci

23:8291–8301.36. Hegg CC, Lucero MT (2004) J Neurophysiol 91:1492–1499.

Czesnik et al. PNAS � February 20, 2007 � vol. 104 � no. 8 � 2971

NEU

ROSC

IEN

CE

Dow

nloa

ded

by g

uest

on

June

14,

202

1

37. Hardy AB, Aioun J, Baly C, Julliard KA, Caillol M, Salesse R, Duchamp- ViretP (2005) Endocrinology 146:4042–4053.

38. Apfelbaum A, Perrut A, Chaput M (2005) Regul Pept 129:49–61.39. Caillol M, Aioun J, Baly C, Persuy MA, Salesse R (2003) Brain Res 960:48–61.40. McLaughlin PJ, Winston K, Swezey L, Wisniecki A, Aberman J, Tardif DJ,

Betz AJ, Ishiwari K, Makriyannis A, Salamone JD (2003) Behav Pharmacol14:583–588.

41. Nieuwkoop PD, Faber J, eds (1994) Normal Table of Xenopus laevis (Daudin)(Garland, New York).

42. Manzini I, Schild D (2003) J Physiol 546:375–385.43. Hamill OP, Marty A, Neher E, Sakmann B, Sigworth FJ (1981) Pflugers Arch

391:85–100.44. Tsou K, Brown S, Saundo-Pena MC, Mackie K, Walker JM (1998) Neuroscience

83:393–411.

2972 � www.pnas.org�cgi�doi�10.1073�pnas.0609067104 Czesnik et al.

Dow

nloa

ded

by g

uest

on

June

14,

202

1